John Sides analyzes them:

The model predicts that these undecided voters will split almost exactly evenly: 50.1% for Obama and 49.9% for Romney. There is substantial uncertainty in this estimate, naturally. The 95% confidence interval for Obama’s predicted vote share is 44% to 56%.

He concludes that "how the undecideds 'break' may not be consequential in this election." Nate Cohn is focused on turnout:

According to national polls, Obama is performing four points better among registered voters than likely voters. That’s well above the more typical 1 or 2 point gap and the main culprit appears to be strong Republican enthusiasm combined with low enthusiasm among young, Latino, and Democratic-leaning independent voters. Since Obama’s coalition is unusually dependent on low-frequency voters, Obama has more to gain from a strong turnout operation than previous candidates. Although it’s unclear whether Obama’s vaunted ground operation can rejuvenate turnout among infrequent Obama '08 voters, the difference between a modest and high turnout among young and minority Obama supporters could easily decide the election. And it's not just that turnout is important, it's that Obama's larger advantage among registered voters makes it an open question whether Obama could actually lose if minority and youth turnout rates approach '08 levels, even if undecided voters broke in Romney's direction.